Simulations of the effect of snow on precipitation

In order to simulate the influences of the TP snow anomalies on the re gional climate in China, the second version of the NCAR (National Center for Atmospheric Research) regional climate model (RegCM2) is used with the model domain containing the TP and the horizontal resolution being 120 km with 11 levels in the vertical. The turbulence kinetic energy (TKE) scheme and a new computational scheme of boundary layer tops with much better effectiveness of simulation are incorporated to replace the original ones in the model (Zheng et al., 1999). The following six experimental schemes are designed.

1. The control experiment (hereafter termed CN). In the CN, the NCEP/NCAR reanalysis 10 year monthly mean fields of meteorological variables, snow depth and snow cover in January from 1986 to 1995 are used as the model's initial data. The time integration starts from 1 January 00:00UTC and ends at 31 August 24:00UTC for 8 months. The nesting values of boundary variables and the sea surface temperatures (SSTs) are also taken from the same period using monthly mean data, interpolated to each day and changed day by day. By using the 10 year monthly mean fields the influences of the background large-scale circulation conditions are removed and the effects of snow are easier to discuss and compare. The model outputs are stored every day; in particular, the model-produced snow depth and snow cover fields are stored every 12 h for later use in other experiments. The purpose of the CN is to check the model performance in simulating the regional climate features and its results are taken as the climatological base for comparisons with other sensitivity experiments.

2. The deeper February snow depth experiment (hereafter termed DL). In the DL, every 12 h in February the snow depth is set to be deeper than that in the CN by a factor of 20~50%, but the snow area is the same as in the CN. Owing to the fact that in the eastern TP the snow depth anomalies have the most evident impacts on precipitation, the snow depth anomalies there are assumed to be the deepest. The other variables are the same as in the CN. By comparing the DL results with those in the CN the effect of the positive winter snow depth anomaly on spring and summer precipitation will be found.

3. The deeper March snow depth experiment (hereafter termed DL2). The DL2 is the same as the DL, but the forcing due to the snow depth anomaly is added in March. By comparing the DL2 with the CN we may find the effect of the positive spring snow depth anomaly on spring and summer precipitation.

4. The shallower February snow depth experiment (hereafter termed DS). In the DS, every 12 h in February the snow depth is set to be shallower than that in the CN by a factor of 20~50% at different grid points over the

TP. The spatial pattern of the snow depth anomaly is the same as in the DL but with about 25% of the amount. By comparing the DS with the CN we may find the effect of the negative snow depth anomaly in winter on precipitation in spring and summer.

5. The larger February snow cover experiment (hereafter termed CL). In the CL, every 12 h in February the snow cover is forced to be larger than that in the CN by expanding the CN-produced snow cover boundary one grid outwards. The other variables are the same as in the DL. By comparing the CL with the CN we may find the effect of the positive snow cover anomaly in winter on precipitation in spring and summer. Then by comparing the CL with the DL, we may find the relative importance of snow cover and snow depth anomalies.

6. The boundary forcing experiment (hereafter termed BN). The BN is designed to study the effect of the background circulation condition at the model boundaries on the interior regional climate and to compare the relative functions of the snow anomaly and the boundary forcing. The boundary variables are taken from the observed ones in the South China Sea Monsoon Experiment (SCSMEX) from 1 May to 31 August in 1998.

In February (Figure not shown), it is clearly seen the simulated snow in the CN occur mainly over the Tianshan Mountains and TP with heavy snow concentrating in the peripheries of TP, particularly in the eastern and western Plateau. Such an essential distribution is fairly coincident with observations (Fig.6.5). In May, the snow mostly melts away due to the temperature rises except over the Tianshan Mountains and the southeastern plateau. It is also shows that the largest time variation of snow and snow depth anomaly take place in the east Plateau.

The peak value of snow depth in the CN appears in February both in the west and the east parts of the plateau. In the DL, it appears in March with a delay of one more month compared to the CN. From the beginning of February the difference in snow depth between the DL and the CN becomes larger and larger, reaching a maximum at the end of the month. From early March the difference gets smaller and smaller and disappears finally at the end of May. The above features of time variations are basically in agreement with the observed ones. The simulated area mean snow depth in the CN is not very much different in both the west and the east of the plateau, while in the DL the snow is much deeper in the east than in the west owing to the different percentage of incorporated snow forcing in the two areas. Therefore the forcing of snow depth anomaly not only influences the difference of snow depth between the two experiments, but also increases the difference of snow depth between the east and the west parts of the plateau, which may enhance the thermal contrast between the west and the east of the TP and therefore influence circulation patterns over the plateau.

Fig.6.7a and b are the differences of precipitation (in mm day - !) between the DL and the CN in May and in summer. It is seen that due to the increased snow depth over the TP in February, the precipitation in May is remarkably reduced in the region south of the Yangtze River. The precipitation in summer is increased roughly between 30°N and 40°N while decreased both south and north of the area. So, the increase of winter snow depth reduces precipitation in the South China area both in May and in summer as well as in the north part of China in summer, while it increases precipitation in the mid and the lower Yangtze River basins in summer. Precipitation over the Bay of Bengal is even more severely influenced by the snow depth anomaly and largely reduced when the snow depth over the plateau increases in winter. Fig.6.7c and d are the same as Fig.6.7a and b but for the BN case. It can be found that the basic patterns of precipitation differences between the BN and the CN are similar to those between the DL and the CN but with much bigger values. The negative anomalies in May in the south of the Yangtze River and South China are more evident and the domain is much larger. The positive anomalies in summer occupy a much larger area between 30°N and 40°N, so that the whole region between the Yangtze River and the Huaihe River becomes an area with more precipitation. Therefore, the background circulation reflected in the boundary forcing conditions has an in-phase effect with the deeper winter snow, though it may be more important to the summer precipitation anomaly. Fig.6.7e and 6.7f show the precipitation differences between the CL and the CN. It is found that the spatial patterns of the differences in May and in summer are somewhat similar to those between the DL and the CN. However, precipitation anomalies are much smaller both in May and in summer except for the Bay of Bengal area in summer where the anomalies are as large as those in Fig.6.7b and even larger than those in Fig.6.7d. Therefore, it may be inferred that the anomalies of snow depth and snow cover are equally important for the summer precipitation in the Bay of Bengal area and more important than that of the background conditions. However, the snow cover anomaly is relatively less important than that of snow depth for the floods over the Yangtze River and the Huaihe River basins in summer. Moreover, the comparison of Fig. 6.7 with the previous result of SVD analysis indicates the similarity of the anomaly patterns of summer precipitation in all three experiments to those in the SVD analysis, though with some discrepancies in area and value.

Fig. 6.7 Differences of precipitation rate (in mm day -1) between the DL and the CN in May (a) and in summer (b); (c) and (d) the same but between the BN and the CN; (e) and (f) between the CL and the CN. (Qian et al. 2003)

The surface energy budget equation can be written as follows (Vernekar et al., 1995):

0 0

Post a comment